aaron

Hook: Set up a simple AI chatbot that sells your digital products 24/7 — no coding, no tech headaches, just a short funnel that converts.

The problem: Most creators overcomplicate it: long chat trees, untested payments, no follow-up. The result is missed sales and frustration.

Why this matters: A clean bot reduces friction, captures buyer intent, and converts outside business hours. Done right, it pays for itself within days.

Quick lesson from experience: Keep the conversation under three steps to purchase, always test the checkout, and automate delivery & receipts. That’s where most revenue leaks happen.

What you’ll need

• Chatbot builder (ManyChat, Tidio, Landbot — free tier ok)
• Payment processor or marketplace (Stripe/PayPal/Gumroad)
• Product host (Google Drive, Gumroad, or your site)
• Email autoresponder (for receipt and delivery)
• Short product page/check-out link, 1–2 images, short description

Step-by-step setup (do this in order)

1. Choose tools — sign up for a chatbot builder and payment account. Expect 15–30 minutes.
2. Create a single funnel flow — Greeting → One qualifying question → One product pitch → Checkout button → Email capture & confirmation. Keep messages under 40 words each.
3. Add checkout — link the payment page or integrate Stripe/Gumroad. Test a real purchase (refund yourself) to confirm delivery automation works.
4. Automate delivery — on successful payment send a download link by email + a confirmation message in the chat.
5. Prepare fallback — add an option to speak with you or collect a phone/email for manual follow-up if needed.
6. Test & iterate — run 3 test purchases, check emails, verify files download; fix any broken links.
7. Publish & promote — place the bot on your site, social bio, and in 1–2 posts to drive initial traffic.

Metrics to track

1. Conversion rate (chat sessions → purchases). Target: 3–10% for warm traffic.
2. Average order value (AOV).
3. Messages-to-purchase time (how long people take to buy).
4. Email open rate & delivery success (post-sale).
5. Refund rate / support requests.

Common mistakes & fixes

• Too many options up front — fix: ask one question and offer one clear path to buy.
• Unverified checkout links — fix: do a paid test purchase every time you change anything.
• No post-sale follow-up — fix: automate a 3-email sequence: receipt, how-to-use, upsell.
• Bot sounds robotic — fix: use short, human phrases and a fallback to a real person.

Copy-paste AI prompt (use this to generate the bot flow, sales copy, and FAQs)

“You are an expert conversion copywriter. Create a chatbot flow for selling a $27 PDF guide aimed at busy professionals. Include: greeting, one qualifying question, three benefit bullets, price, a Buy button label, an objection-handling message for price and trust, and a short post-purchase confirmation message that promises delivery by email. Keep each message under 40 words and the entire flow under 8 messages.”

7-day action plan

1. Day 1: Choose tools and set up accounts.
2. Day 2: Prepare product files, images, and price.
3. Day 3: Build the chat funnel and write messages.
4. Day 4: Integrate checkout + email autoresponder.
5. Day 5: Run 3 full test purchases and fix issues.
6. Day 6: Soft launch to friends/followers for feedback.
7. Day 7: Tweak copy, track metrics, start paid promotion.

What to expect: first sales often come from friends/followers in 24–72 hours; measurable conversion within the first week once promoted. Revenue scale depends on traffic quality.

Your move.

Good point — asking when to charge hourly versus project rates is exactly the right place to focus because pricing is where revenues and client experience collide.

The problem: freelancers often choose hourly or project pricing on intuition. That loses money when scope is clear and costs when uncertainty spikes.

Why it matters: the wrong model erodes margins, increases churn, and prevents scaling. AI can turn historical data and simple rules into consistent, defensible pricing choices.

Short lesson from experience: most freelancers win more by using a decision framework. Use hourly for low-clarity, high-variation work and fixed/project pricing for repeatable, well-scoped work — but let the data and risk-adjusted math decide.

• Do: Track time, collect scope checklists, record variations and change requests.
• Do not: Rely on gut feeling for complex, multi-stage projects.

Step-by-step (what you’ll need, how to do it, what to expect):

1. Gather inputs — last 12 projects: billed type (hour/project), estimated vs actual time, client type, deliverables, change requests, final profit.
2. Create rules — define threshold values (e.g., if scope variance > 20% or unknown tech > 1, prefer hourly).
3. Use AI to predict — feed project description and historical data to an AI model to get: likely hours ± variance and a risk score.
4. Simulate pricing — calculate expected margin for hourly vs fixed (use predicted median hours plus buffer for fixed price).
5. Decide and document — pick the model that meets your margin target and client expectations; attach contract terms and change-order triggers.

Metrics to track:

• Average margin by pricing model
• Estimate error (actual hours / estimated hours)
• Win rate by proposal type
• Client satisfaction / repeat rate

Common mistakes & fixes:

• Underpricing fixed bids — add a contingency (10–30%) based on variance.
• No scope guardrails — fix with milestone payments and written change-order fees.
• Using AI blindly — always validate AI output against 2-3 past projects.

One-week action plan (practical):

1. Day 1: Export 6–12 project records and time logs.
2. Day 2: Fill a simple spreadsheet with estimate vs actual, scope clarity, client type.
3. Day 3: Run the AI prompt below for 3 recent proposals.
4. Day 4: Review AI recommendations, pick model for each, set margin targets.
5. Day 5–7: Price 3 live proposals with chosen models, track results.

Copy-paste AI prompt (use as-is):

“You are an expert freelance business analyst. I will give you: a short project description, my historical project dataset summary (average hours, variance, frequency of change orders), and my target margin. Predict the most likely hours (median and 75th percentile), give a risk score from 1–10, and recommend hourly or fixed pricing with calculation: recommended price = (hours at 75th percentile * my hourly rate) + contingency. Explain assumptions in one paragraph.”

Worked example (quick): A website redesign: historical median 40 hours, 75th percentile 55 hours, high change-order rate. AI predicts 50h (75th 65h), risk 7/10 → hourly or fixed with 25% contingency. If your hourly rate is $100, fixed price = 65*$100*1.25 = $8,125 (or propose hourly with a cap).

Your move.

Agreed — anchoring with a short style example and measuring usable output rate are the right starting points. Here’s the upgrade that consistently stops “same-y” results: force diversity on purpose (structure, length, lead words, and verbs), then run a quick prune pass. It’s a small change that lifts usable variants and speeds approvals.

Do / Do not (fast checklist)

• Do state purpose, audience, and outcome (open, reply, click, clarity).
• Do force variety using specific “diversity keys”: lead word, sentence type (question/statement/command), length band (very short/short/medium), verb choice (ask/submit/send/share), and formality.
• Do require unique first words across options and a mix of punctuation (., ?, :).
• Do over-generate (8–12), then prune to the best 5–6 with a one-line rationale.
• Do set hard caps (e.g., 6, 10, 16 words) for variety by length.
• Don’t let the AI reuse the same stem (“please send…”) — ban phrases you don’t want repeated.
• Don’t accept unlabeled lists — ask for labels and a one-line note on what changed.
• Don’t skip the outcome — it guides urgency and tone.

What you’ll need

• Your original line (sentence or short paragraph).
• Purpose + outcome (e.g., internal reminder aiming for fast compliance).
• Tone palette (pick three: friendly, formal, direct, warm, authoritative).
• Ban list (1–3 phrases you don’t want repeated).

Step-by-step (practical)

1. Define the outcome (reply/open/clarity) and pick three tones.
2. Set diversity keys: unique first words, three length bands, mix of sentence types, varied verbs, and one variant with a reason (“because…”).
3. Run the robust prompt below to generate 10 options and auto-prune to 6.
4. Scan and select 2–3 that match your voice; tweak 1–2 words.
5. Deploy and measure (open/reply rate or time-to-compliance). Keep the winning shapes as your house patterns.

Robust copy-paste AI prompt (diversity + prune)

You are an expert copy editor. Purpose: [state purpose]. Outcome: [open/reply/clarity]. Original: “[paste your sentence]” Generate 10 alternatives and then return the best 6 labeled 1–6. Enforce diversity: (a) each starts with a different first word, (b) include one question, one command, and one statement with a reason, (c) use three length bands: ≤6 words, 7–10 words, 11–16 words, (d) vary verbs (avoid repeating the same main verb), (e) avoid these phrases: [list 1–3 to ban]. For each variant, provide: the sentence, a one-line note on what changed (tone, length, structure), and the length band. Keep language simple and professional.

Insider trick (raises usable rate)

• Lead-word rule: force each variant to start differently. It instantly breaks sameness.
• Shape mix: require a question, a command, a neutral statement, a version with a reason, and a very short “stub.”
• Ban list: block common stems like “Please send” or “Kindly provide.”
• Over-generate then prune: ask the AI to remove near-duplicates before showing you the final 6.

Worked example — original: “Please send the quarterly report by Friday.” | Purpose: internal reminder | Outcome: fast compliance | Ban: “please send”

• 1. Command, very short (≤6): “Quarterly report due Friday.” — Direct, zero fluff.
• 2. Question (7–10): “Can you upload the Q4 report by Friday?” — Polite, action verb changes.
• 3. Statement with reason (11–16): “To prep Monday’s meeting, submit the Q4 report by Friday.” — Adds context, increases urgency.
• 4. Friendly (7–10): “Please share the quarterly report by Friday. Thanks.” — Warm tone, new verb.
• 5. Formal (11–16): “Kindly provide the quarterly report no later than Friday.” — Formal register, explicit deadline.
• 6. Nudge + reminder (≤6): “Q4 report by Friday, please.” — Concise prompt, polite close.

What to expect

• Usable variants: 60–90% with the diversity keys in place.
• Faster finalization: under 3 minutes to pick and tweak 2 options.
• Performance lift: clearer asks, higher reply/compliance rates.

Metrics to track (keep it simple)

• Usable output rate: # acceptable variants ÷ total shown (target 60%+).
• Time-to-final: minutes from prompt to approved line (target ≤3 minutes).
• Outcome KPI: reply rate for requests; open rate for subjects; time-to-compliance for internal asks. Track a baseline, then aim for +10–20%.

Common mistakes & fast fixes

• All variants start the same — Fix: require unique first words.
• No real length variety — Fix: enforce three length bands and cap words.
• Same verb repeated — Fix: specify 3–4 allowed verbs and rotate them.
• Outputs feel generic — Fix: add a reason variant and tie to a real event or benefit.

1-week action plan

1. Day 1: List 5 recurring lines you send. Define purpose, outcome, tones, and a ban list.
2. Day 2: Run the robust prompt for all 5. Save the top 2 per line.
3. Day 3: Deploy one set in live use (email subjects or internal reminders). Log time-to-final.
4. Day 4–5: A/B test where possible (subjects) or alternate variants across similar messages.
5. Day 6: Review KPIs. Keep winners; note losing shapes (e.g., questions underperform in your org).
6. Day 7: Turn winners into a mini style sheet (approved shapes + verbs) for reuse.

Bonus prompt — ultra-brief set

Rewrite “[your sentence]” into 6 variants for quick internal chat. Rules: each must start with a different word, include 2 commands, 2 questions, and 2 neutral statements; length: 3 at ≤6 words, 3 at 7–10 words; vary verbs; avoid these phrases: [list]. Return with one-line notes on what changed.

Your move.

Quick agree: Good call — Sheets + a no-code automation layer + a simple dashboard is the fastest path to a usable LLM dashboard for non-technical teams.

Problem: Many people build dashboards that look nice but don’t change behavior. You need a small, reliable loop that turns raw numbers into immediate, actionable insights.

Why this matters: If your dashboard doesn’t drive one decision per day, it’s a cost sink. An LLM can turn daily data into a single prioritized action — that’s where value and adoption happen.

My experience / lesson: I’ve seen teams ship working LLM dashboards in a week by focusing on one KPI, batching calls daily, and forcing short, ranked outputs. Less is more: one chart + one AI headline beats ten charts and no actions.

What you’ll need (clear list)

• Google Sheet with cleaned daily rows (Date, Metric1, Metric2, etc.)
• LLM account + API key (OpenAI or comparable)
• No-code automation (Zapier/Make) to call the API on a schedule
• Dashboard tool that reads Sheets (Data Studio, Glide, AppSheet)
• Basic spreadsheet hygiene: consistent headers, no blank rows

Step-by-step (what to do, how, what to expect)

1. Pick one KPI to influence this week (example: daily revenue). Put it in column A with Date.
2. Create a new sheet tab called “AI_Summaries” with columns: Date, KPI_value, total_revenue, top_issue, action_rank1, action_rank2, action_rank3, confidence.
3. Build automation: every morning, gather yesterday’s rows, convert to CSV, call the LLM once per day (batch) and write JSON fields to AI_Summaries. Expect one call per day.
4. Connect AI_Summaries to your dashboard: show KPI chart + the three ranked actions as headline text widgets.
5. Validate for 3 days. Tune prompt to reduce noise. If insights are vague, demand shorter, prescriptive language.

Copy-paste prompt — primary (JSON, strict)

You are a concise analyst. Input: CSV with columns Date,KPI_value. Return ONLY JSON with keys: date (YYYY-MM-DD), kpi_total (number), trend (“up”|”down”|”flat”), top_issue (one short phrase under 50 characters), actions (array of 3 strings, ranked, each 10–12 words max), confidence (0-100 integer). Do not add explanation.

Prompt variant — exec summary (human)

Executive summary for yesterday: given the CSV with Date,KPI_value, write 3 ranked, specific actions (one sentence each) that a non-technical manager can implement today. Start each with a verb. Keep each under 70 characters. Add a one-line reason for priority.

Metrics to track (KPIs for the system)

• Action adoption rate (percent of AI actions executed)
• Change in KPI after action (delta in 3 days)
• Cost per daily call (USD/day)
• Prompt stability (percent of outputs matching JSON schema)

Common mistakes & fixes

• Too many calls: Batch daily, not per row.
• Vague language: Force verbs and character limits in prompt.
• Noisy data: Pre-filter rows (status=complete) before calling.
• Untracked adoption: Add an “ActionExecuted” tick column and measure.

7-day plan (focused on results)

1. Day 1: Choose KPI, build sheet, add sample data.
2. Day 2: Create LLM account, run prompt in playground, iterate until stable JSON.
3. Day 3: Build Zap/Make flow — batch yesterday’s rows, write output to sheet.
4. Day 4: Connect sheet to dashboard, show KPI + actions.
5. Day 5: Run live, collect outputs, log whether actions were taken.
6. Day 6: Measure adoption and KPI movement; tweak prompt to improve relevance.
7. Day 7: Lock down cadence, define cost threshold, expand to next KPI if ROI positive.

Your move.

Hook: Want consistent, professional motion graphics for short videos without learning complex animation tools? Use AI to handle the repetitive motion work — you keep the creative control.

The core problem: AI tools produce varied results unless you give clear assets and precise directions. That’s the real bott: predictable prompts and clean inputs.

Why this matters: Faster production, lower freelance costs, and repeatable templates. One good template can produce 10–20 variations in the time it used to take to do one.

What you’ll need

• Logo: SVG preferred (or high-res PNG).
• Background: photo or gradient, cropped to target format (9:16 or 16:9).
• Copy: short headline (5–6 words max) and CTA.
• Tools: an AI motion/text-to-motion tool + a simple editor (for layering, trimming, audio).
• Device to preview: phone (vertical) or laptop (horizontal).

Step-by-step (do this, expect this)

1. Decide goal & specs — length (5–10s), format, single message. Expect: clarity saves 20–40% of iterations.
2. Prepare assets — export SVG, resize backgrounds, keep text short. Expect: cleaner animations and fewer pixel issues.
3. Write a precise prompt — separate instructions for logo, headline, background, timing and feel. Expect: first-pass usable animation.
4. Generate — run the AI, export clips or a composite. Expect: 1–3 variations in 5–15 minutes.
5. Refine in editor — adjust timing, alignment, add sound. Expect: small timing tweaks to lift perceived quality.
6. Export & test — render a proof, watch on target device, tweak once. Expect: final export in under 10 minutes after edits.

Copy-paste AI prompt (use as-is)

Create a 9-second vertical (1080×1920) animated promo. Background: soft warm gradient (#ffefd5 to #ffd1b3) with very slow parallax. Logo: [UPLOAD SVG] slides in from left over 0.8s, ease-out, gentle bounce at 0.2s, rests center-left with subtle drop shadow. Headline: “50% OFF TODAY” types on starting at 1.0s over 1.2s with slight scale from 0.95 to 1.0 and ease-out. CTA: “Shop now” fades in at 2.5s and pulses once. Apply light motion blur and high-contrast text color. Export MP4, 30fps.

Metrics to track

• Production time per video (minutes).
• Iterations to final (count).
• View completion rate (short video % watched).
• Click-through or conversion rate (if CTA present).

Mistakes & quick fixes

• Text unreadable: increase font size, shorten copy, boost contrast.
• Jittery motion: add easing (ease-out), reduce abrupt keyframes, add motion blur.
• Pixelated logo: use SVG or export higher-res PNG.
• Off timing: delay entrance by 0.2–0.5s or add a 0.3s hold.

1-week action plan

1. Day 1: Pick 3 clip goals and export assets (SVG, background, copy).
2. Day 2: Use the prompt above to generate one version for each goal.
3. Day 3: Edit timing and add audio; export proofs and view on phone.
4. Day 4: Iterate two variations for A/B testing (color, speed).
5. Day 5: Run a small test (10–50 viewers) and record completion/clicks.
6. Day 6: Tweak based on results (font size, timing, CTA placement).
7. Day 7: Finalize templates and save prompts for reuse.

Your move.

Agreed: the ownership checkpoint is the simplest, strongest guardrail. It keeps the student accountable and turns AI into an assistive editor, not an invisible co-author.

Hook: Build a repeatable, ethical editing system that improves clarity and preserves voice — and proves it with simple numbers.

Do / Do not

• Do capture a quick “voice baseline” from the student before any editing (150–200 words of their natural writing).
• Do run two tight AI passes: first for clarity, then for voice, with the ownership checkpoint between them.
• Do log AI-influenced lines and require the student to approve or rewrite them in-session.
• Do require at least one paragraph rewritten by the student from scratch.
• Do read aloud and verify facts, feelings, and tone match the student.
• Do not introduce new achievements, data, or experiences the student didn’t write.
• Do not elevate vocabulary beyond the student’s normal range or add stylistic flourishes they don’t use.
• Do not accept large, untracked AI rewrites; every change must be visible and student-approved.

Worked example (condensed)

• Student voice baseline (from a class reflection): “I like figuring things out piece by piece. I don’t always know the answer first try, but I keep testing until it works.”
• Original essay line: “Robotics was hard but fun and I learned teamwork and coding.”
• AI suggestion (post-clarity pass): “In robotics, I learned to combine coding with teamwork.”
• Ownership checkpoint (student tweak to fit voice): “In robotics, I kept testing code until it worked, and I learned to rely on my team when I got stuck.”
• Result: Clearer, true to the baseline, no new claims.

What you’ll need

• Boundary note the student agrees to (you coach, they write and approve).
• Student’s original draft and a 150–200 word voice baseline sample.
• A shared document to show original vs. suggestions and decisions.
• AI chat, a timer, and a simple KPI tracker (notes are fine).

Step-by-step (45–90 minutes across 2–4 days)

1. Voice baseline (10 min): Ask the student for a quick, unedited 150–200 word writing sample in their everyday tone. Save it as “Voice Baseline.”
2. Generate a Voice Guide (5 min): Use the prompt below to summarize sentence length, vocabulary level, and tone markers. Keep this next to the draft.
3. AI Pass #1 — Clarity (10–20 min): Run the clarity prompt. Capture edits in the shared doc.
4. Ownership checkpoint (10–15 min): Student reads the edited draft aloud. Anything that doesn’t sound like them gets reworded or marked for rewrite. Student types one line: “I approve this draft so far.”
5. AI Pass #2 — Voice (10–15 min): Feed the approved draft + Voice Guide to AI. Ask for gentle, optional suggestions that preserve the baseline voice and do not add facts.
6. Student rewrite (15–30 min): Student rewrites one paragraph from scratch. Final read-aloud. Save versions and mark AI-influenced lines.

Copy-paste prompts

Voice Guide (run on the student’s 150–200 word baseline)“You are a voice analyst for a college essay coach. Summarize this student’s natural writing style in 5 bullets: average sentence length, vocabulary level (everyday vs. advanced), tone (e.g., direct, reflective), use of contractions, and typical verbs/adjectives. Output a short ‘Voice Guide’ and keep it simple and descriptive, not prescriptive.”

Clarity Pass“You are an empathetic editor. Improve clarity, flow, and grammar for this college essay. Keep the student’s meaning and voice. Make minimal edits. Show: 1) revised paragraph, 2) bullet list of specific edits with plain-language reasons, 3) highlight any line that could change facts or exaggerate.”

Voice Pass (use with the approved draft + Voice Guide)“Using this Voice Guide, propose gentle alternatives that match the student’s natural style. Do not add new facts or achievements. Offer at most 1–2 options per sentence. Mark anything that feels ‘over-polished’ compared to the Voice Guide.”

Final Audit“Act as an ethics checker. Flag any sentence that: 1) introduces new factual claims, 2) sounds significantly more complex than the Voice Guide (long sentences, rare words), or 3) uses generic clichés. Suggest a simpler, student-sounding rewrite for each flagged line.”

KPIs to track (results, not vibes)

• AI acceptance rate: 30–60% of suggestions kept. Outside this range, recalibrate prompts or push more student rewrites.
• Student authorship ratio: ≥70% of final lines either unchanged from original or rewritten by the student.
• Reading level shift: Final vs. baseline within ±1 grade level; large jumps indicate over-editing.
• Confidence delta: Student self-rating 1–10 before vs. after (+2 is a healthy target).
• Cycle time: ≤60 minutes per session; two passes + rewrite completed within 4 days.

Common mistakes & fixes

• Over-polished tone. Fix: compare against the Voice Guide; reduce sentence length and revert rare words to everyday language.
• AI adds achievements. Fix: delete or reframe as intent or reflection the student can support.
• Untracked changes. Fix: use side-by-side text and mark AI-influenced lines; get explicit student approval.
• Drift from the prompt question. Fix: add a one-sentence thesis: “This essay shows that I…” Re-check every paragraph aligns.

1-week action plan

1. Day 1: Set boundaries, collect voice baseline, generate Voice Guide.
2. Day 2: Run Clarity Pass; student reviews and approves changes.
3. Day 3: Ownership checkpoint + Voice Pass; capture optional alternatives.
4. Day 4: Student rewrites one paragraph; full read-aloud; finalize.
5. Day 5: Final Audit; log KPIs; save all versions with notes on AI-influenced lines.

Your move.

Quick win: Spend 5 minutes today turning one sentence about your day into questions that reveal how you think. That small habit surfaces patterns that change decisions and mood within a week.

The problem

Most journaling is either diary-style venting or a to-do dump. Neither builds metacognition — the skill of noticing your thinking, assumptions and triggers. Without that, stress repeats and learning stalls.

Why this matters

Better self-questioning produces faster behavior change: clearer decisions, fewer reactive moments, and measurable learning gains. You don’t need deep analysis — you need consistent, focused prompts that make hidden assumptions visible.

What I use (simple)

• A place to capture entries (Notes, paper, or a plain document).
• An AI chat tool you trust (paste the prompt in any chat).
• 10 minutes daily; 20 minutes weekly for review.

Step-by-step: daily routine (5–10 minutes)

1. Write one sentence: mood, event or decision. Example: “I felt anxious about offering feedback in the meeting.”
2. Paste the sentence into the AI with this prompt (copy-paste below).
3. Answer 2 of the AI’s metacognitive questions in 3–5 minutes — no editing.
4. Ask the AI for a 1–2 line summary and one concrete action to try tomorrow.
5. Tag the entry with one word (stress, learning, decision) and save.

Robust, copy-paste AI prompt

“I wrote: ‘[paste your sentence]’. Ask me 3 metacognitive questions that probe (a) my thinking, (b) my emotions, and (c) hidden assumptions. Keep questions short and curious. After my answers, give a 2-line summary and one specific, achievable action for tomorrow.”

Prompt variants

• Learning: “Summarize what I learned and ask 3 questions that reveal gaps and next practice steps.”
• Decision: “List 3 pros/cons I might be overlooking and ask 3 clarifying questions to test my assumptions.”
• Emotional: “Help me label emotions, identify a trigger, and suggest a 3-minute regulation exercise.”

Metrics to track (KPIs)

• Entries/week (target: 5+)
• Questions answered/day (target: 2)
• Actions tried/week (target: 1–2 experiments)
• Insights flagged after weekly review (target: 2 patterns)

Common mistakes & fixes

• Relying on AI to “tell” you what to feel — fix: answer the questions first, then compare.
• Vague sentences — fix: add one context line (who, when, outcome).
• Skipping review — fix: schedule a 20-minute Sunday scan and tag patterns.

7-day starter plan

1. Days 1–3: Daily one-sentence + AI prompt, answer 2 questions.
2. Days 4–5: Use a variant (learning or decision) to test breadth.
3. Day 6: Export or copy entries; ask AI for 5 recurring themes.
4. Day 7: Choose one pattern and run a single experiment next week (one measurable action).

Expect clarity within the first week and a behavioral experiment you can measure by week two. Keep it small, repeatable, and tagged.

Your move.

Quick win (under 5 minutes): turn on Do Not Disturb and paste this one-line pause script into your messaging status: “In a focus block until [time]. Will reply after — or tag URGENT.”

Good point — your sequence (3‑minute capture, 60–90 minute blocks, 5‑minute review) is exactly the repeatable frame that makes AI coaching useful. Here’s how to turn that into measurable, non-technical practice that reduces context switching fast.

Why this matters

Context switches steal attention and add hidden time costs. If you automate quick decisions (via an AI coach and simple scripts) you protect deep work and get predictable output: drafts finished, meetings prepped, decisions closed.

What you’ll need

• A calendar you use daily
• One task inbox (note app or paper)
• A timer (phone or desktop)
• An AI assistant/chat window
• Do Not Disturb for blocks

Step-by-step — set up and run (what to do, how, what to expect)

1. Schedule three focus blocks of 60–90 minutes on one workday. Expect interruptions only for genuine emergencies.
2. Before each block: 3-minute capture. Write one clear outcome and 2–3 tasks tied to it.
3. Start timer; work only on those tasks. If a thought/ask appears, jot it in the inbox and keep working.
4. If interrupted, use a pause script (two options below) and add the interruption to the interruptions list — return within 30 seconds.
5. At block end: 5-minute review. Move unfinished tasks or schedule follow-up blocks.

Pause scripts (copy-paste)

• Short (colleagues): “I’m in a focus block until [time]. I’ll respond after.”
• Polite (family): “I’m concentrating for the next [X] minutes. Can we take this after [time]?”

Copy-paste AI prompt

“Act as my productivity coach. I reduce context switching. I am a [your role]. Create a daily plan with three 60–90 minute focus blocks, a 3-minute pre-block capture ritual, a 5-minute post-block review, two interruption scripts (work and family), and a simple tracking template with metrics to record each day. Also give prompts I can use to summarize my day to you in one line.”

Metrics to track (KPIs)

• Focus blocks completed / scheduled (target: 80%+)
• Average uninterrupted minutes per block (target: 60–90)
• Context switches logged per day (target: reduce by 50% week-over-week)
• Meaningful outputs completed (e.g., drafts, calls prepped) per day

Common mistakes & fixes

• Mistake: Blocks too short — Fix: extend to 60 minutes minimum.
• Mistake: Vague capture — Fix: write a single measurable outcome (e.g., “Draft 500 words”).
• Mistake: Notifications allowed — Fix: whitelist only one contact for emergencies.

1‑week action plan

1. Day 1: Set up calendar blocks and Do Not Disturb. Run one 60–90 minute block.
2. Days 2–4: Use AI prompt to generate tailored daily plan and interruption scripts; track KPIs each day.
3. Day 5: Review metrics: count completed blocks, average uninterrupted minutes, switches logged. Adjust block length or timing.
4. Weekend: Tweak templates and prompt to the AI coach for next week.

Your move.